The Abstraction Stepladder - Part 1
How Much Can We Level Up?
Have you ever taken the time to think carefully and deeply about what Abstraction is?
By Abstraction, I mean the ability to create a concept or representation that generalizes the data we see. In effect, it puts a handle on “things” that don’t have handles.
I have hesitated for months to write this important essay because I was worried that people might find it, well, abstract. But let me assure you, we will be using many concrete examples.
The first example is the adjective “concrete.” You know what I mean by it…you have seen a lot of concrete in your life, and it’s hard and solid and reliable. So, your brain has taken that data and created a concept of “concreteness.” It’s like you took a series of properties associated with concrete and wrapped them up and put a label on it, and it has a handle sticking out of that package (the word “concrete” itself) that you can use to lug that concept around. You can then pull that meaning-package out at will and slam it onto some other abstract concept (like the word “example”), and then that second concept has been solidly imbued with the list of properties that remind you, metaphorically, of real concrete.
Obviously, we do this Abstraction thing all the time. Humans are amazing at it. We are going to explore the insane levels of power that humans can obtain using it. But first, what is Abstraction, and how does it work?
Well, we sort of know how it works. Data comes into the human brain, say from our eyes. Let’s say that we are looking at a concrete sidewalk with a big red ball on it.
A 2D image of colors is transmitted from our retinas up our optical nerve to our visual center in our brain. The data is parsed and sent to our neocortex. Then different modules of neural networks bring that image up through themselves, performing data processing.
First, color detection happens - we apprehend the average color of the center of what we are looking at. A signal is sent that we need to look at it…it’s bright red! We are programmed to notice red as a danger signal (maybe because blood is red?).
Second, edge detection happens - we apprehend that there is a circle. It has a shadow. There is a white glare area in the upper right. The ground has converging almost parallel lines at its edge, but on the right there are a lot of parallel lines close together. Stereoscopic vision (from our two eyes) is probably incorporated together at this point somewhat, so some depth is assessed.
Third, relative changes in color and shading are detected - we apprehend that the red circle is a sphere from the fact that its upper right is way brighter in color than its lower left, and the colors smoothly change. Also, based on other color changes on the sidewalk, we can tell the sphere is resting on and touching the sidewalk. We might also infer textures of the ball (smooth, but look at the fine color variation) and the sidewalk (rough, but especially so on the right). This third step is clearly building on the “discoveries” of the first two steps.
Fourth, we start to bring old and recent historical data to help know what it is without touching it - we remember that one time we sat on it, so the ball is heavy or fixed. We touched it, and its coldness implied certain materials or massiveness. It’s a concrete ball anchored in place on the sidewalk. And we’ve only seen these in front of the American Target stores. We must be looking at the front of a Target store. So we have built on the third step plus a ton of steps that came before, and come up with a pretty impressive inference - we are in front of a Target store.
The abstractions might stop there, or we might go further and try to assess whether the designers are being playful, or security-minded. Did Target start putting these barriers up before or after 9/11? Does the spacing work as a security barrier? How does a bright red ball make me feel? Make my kids feel? How would the impact on my mood impact my purchasing?
Most animals clearly do the first three steps, and effectively do something very like the fourth step, at least in recognizing familiar places. But human minds have so much extra computational horsepower (mostly in our neocortex) that we can take those cool Abstraction abilities and go much, much further. In “The Simulation In Our Heads,” I discussed how our brain can repurpose our super-fast Social Calculator to create an uber-powerful physics, economic, or whatever simulation in our head. But we can do lots of other things with our brains, too…with the power of Abstraction.
In effect, humans have climbed the Abstraction ladder so high that we have subconsciously Abstracted Abstracting.
Meaning, we can take a highly functional toolkit and use it to perform all new types of jobs. Time for another concrete example:
Humans can be taught to take their edge-finding and fine detail mapping visual capabilities and focus in on twig-like high detail markings to rapidly distinguish different types. Then we can tie the output of sequential recognitions of those tiny markings to our vocal recognition and language processing center. You are actually doing this right now, in reading this paragraph.
Repurposing the ability to Abstract in many different ways has become so fundamental to the modern human experience that most of the time we don’t even realize we are doing it. As I said before, we do it subconsciously. Think about how amazingly powerful humans as a species have become because of reading and writing. We would not be in space, have nukes, or cure hardly any diseases.
The reason writing was a latchkey progression that unlocked so much else is not just that it allowed for the storage of learned information beyond the timeframe of single lifetimes. Writing changed how we think1. Writing down words made the words non-ephemeral. It made them concrete. Because the words sometimes represented ideas, then the ideas also became more concrete. Here’s how Derek Hodgson describes it, in terms of neurophysiology:
“These developments therefore enabled the brain to reuse the visual cortex for an entirely new purpose. Ultimately, it could have created a new process in the brain that exploited the visual cortex, giving rise to a visual word form area and connecting with speech areas incrementally over time.”
Writing allowed concepts to be solidified and codified. And once this codification of an idea was widely accepted, the idea could be transmitted faithfully between people, and argued about, and improved or disproved. The strongest concepts did not just become widely known. They became widely built upon. Entire societal structures could then be constructed that otherwise had no way to come into being.
One of the best examples of this is property.
7,000 years ago in most places, the concept of property did not exist. But we can probably trace how the modern idea of property developed sequentially, via higher and higher levels of abstraction.
The area between the lone oak and the stream was Jarek’s, because Jarek was strong and he chased away anyone that came close to this area. It was not the land of Jarek, but rather the land that Jarek fought to keep.
Jarek’s great grandson Tubal built a fence around the entire land. This was called Tubal’s Fence, because Tubal built the fence. Tubal did not make the stones, but by 6820 B.C. in Sumeria, everyone had abstracted out that a linear pile of stones surrounding an area is called a fence. And the land within the fence is different than the land outside a fence.
Over the next 100 years, a city sprang up next to Tubal’s Fence (as his great-grandchildren called it), and people began writing about trade to keep ledgers. The pomegranates that came from the place of Tubal’s Fence were able to fetch a higher price than those elsewhere, and the writing reflected that. In naming the land “the land of Tubal’s fence,” or just, “Tubal’s land,” the idea of ownership of this land entered the minds of dozens of merchants. They accepted the concept because it was useful and written and widely accepted.
Others saw the value of holding land, and also the value of the rising city-state in formalizing things. Others built their fences around the plots they had been protecting continuously. Maps were drawn with the river and the oak and many other landmarks, criss-crossed by lines representing the fences. The land pieces were assigned on paper in tables, to the people that owned the land. The common agreement was that no one would fight over the land, because all agreed that the fence lines were good, because peace was good, because peace meant stable trade, which kept most people from starving as population density and specialization increased. This was about 3 levels of Abstraction above Jarek yelling threats at trespassers, and it made life a lot better for a lot of people.
In order to provide for the common defense, taxes were levied, including annual tributes of 1/60th of the produce of each land piece. This allowed of the specialization of rulers and their armies. The name of Larak was given to the city and its armies. Notice was given that Larak was fearsome and ruthless to any that attacked, but generous and welcoming to traders. Thus the land of Larak in Sumer flourished in safety and wealth, built on perhaps 5 layers of Abstraction beyond Jarek.
And on this foundation of the concept of property grew kingdoms and states and countries and empires…all higher and higher levels of Abstraction. Entire ethnic groups became named after their land, not vice versa. And each higher level of Abstraction gave those humans more power to shape their social and physical world, for better or for worse.
But then the Abstraction of property did not just go up in hierarchical levels. The concept went sideways. Long ago property began to mean not just land, but also physical holdings, such as the buildings on the land. Then property came to mean the stuff stored in the buildings or the ships that carried the stuff to far away ports. About 8 Abstraction levels above Jarek.
By delineating ownership of large piles of things, along with the assumed right to not have that stuff just taken by someone stronger than you, the basis for other Abstractions condensed out of human minds.
Governments did not just protect from foreign invaders, but also defended property rights of one citizen against the other stealing it. Property Rights were enumerated and internally enforced. Less resources of the individual were devoted to guarding, and more to merchant work, and wealth increased.
The next big Abstraction was paper representations of pieces of land (deeds), which traced ownership via family lines. Eventually, someone realized that with a good system of unique stamps, once could also create a ledger of ownership of stuff. Bills of ownership of flasks of wine and transfer records of that property could be exchanged. Here is Assyrian real estate deed from 7th century B.C., courtesy of the Met Museum.
Then someone else realized those bills of receipt are tradeable themselves, without anyone having to touch wine or barley. A trade of wine or barely could be made without having to physically even move them!
Then someone Abstracted out the idea that a promise of wine could be made using these receipts for barley received right now. Property futures were invented. This allowed, for the first time, the time-shifted application of wealth — investing.
Fast forward past the creation of promissory notes and the first corporations and sovereign bonds. Money came into being, the ultimate fungible asset. This added so much flexibility that everyone grabbed onto the concept once introduced to it. Money, about 14 levels of Abstraction beyond Jarek threatening trespassers with a sharp stick, changed the world. Its attraction is that it allows a human that has it the flexibility to do thousands of different things…way more than a wagon load of barley by itself would allow.
Think of all the institutions that were then build upon the existence of the purely abstract concept of money. Banks and formalized investment, tax regimes, fines for crimes. Granted, not all of these are good. But they are very powerful.
Let’s be very clear here: It’s not that these particular Abstractions are powerful. ALL Abstractions are potentially very powerful, especially as foundations for even higher Abstractions.
I propose a Principle of Abstractive Power: The more abstracted out a concept becomes (while still connected to utility), the more deeply and broadly useful that concept is.2
And we are on the cusp of further Abstractions in this space. Cryptocurrency basically turns electricity into money using math. This is now about 26 levels of Abstraction above Jarek’s pointy stick. And we are not done yet. There are single Non-Fungible-Tokens (NFTs) of 64x64 pixel zombies selling for more than Fabergé eggs. Distributed Autonomous Organizations (DAOs) have just recently popped onto the scene, and their power may soon eclipse some small companies, and perhaps even some small governments in the next few years. Web3 applications may be somewhat overblown, but networks of blockchain Smart Contracts might be used to create investment banks that have zero human employees in the next decade. Abstractions on top of Abstractions on top of Abstractions.
I write all this because for most of us, Abstractions are the air we breathe — a fundamental and ubiquitous part of being human. So much so that we don’t even know we are doing them.
But the levels of difficult Abstraction are not just present in the world of particle physics, higher math, and music theory. They are reaching out to shape the world around us. People had millennia to get used to the concept of property, and centuries to warm to the concept of money. Most of these humans had the luxury of not really thinking about the concepts as Abstractions or the process of Abstraction itself. They were able to just sort of ease into it by osmosis and incorporate it slowly into their lives and societies.
But now many things, not just crypto things, are happening in multi-tiered, Abstract ways very quickly. And so, we must stop Abstracting subconsciously only. We need to think consciously and overtly about Abstraction itself.
Why does cryptocurrency adoption potentially change how business is done?
How could I help poor people via virtual properties, such as in-game items?
Are DAOs the savior or doom of democracy?
Is there a way to algorithmize my greatest talents and scale them up to wider use?
AI is learning to be very persuasive to humans. What will the humans wielding this power do with it? What CAN they do with it?
We cannot intuit our way to mastery of these things. We must develop a framework for analyzing the Abstractions and their power, both in our head and in our society.
The example of property is a chain of Abstractions that were built in society as a whole. But what about you?
As we learn and grow and think, we can carefully climb personally the same Stepladder of Abstractions that society has. In fact, I have spent the last few decades doing exactly this. Climbing that stepladder is one of the main things that allow me to function in my job and write this blog.
I have a relative that clears, levels, and prepares land for a living. He installs drainage systems, lays the foundations for parking lots, and creates precise landscape shapes. Here is a pic from one of his sites.
In order to quote this work and perform it, he has Abstracted an insane amount of knowledge into useful frameworks of how dirt compaction works, the economics of running large diesel machines, the acceptable regional pricing for his work, and how to best handle tree stumps. He’s super effective because he’s abstracted a ton of work experience into principles of problem solving.
You do the same in your life, but at higher or lower levels. We all do this subconsciously.3
What if we started to play the Abstraction game consciously? How do we do so? What if we develop rules and laws and frameworks for creating and applying Abstractions themselves? As best I can tell, no one has ever formalized something like this.
I didn’t write this article because I know how to do it. I don’t, yet. But I wanted to make the issue concrete, so maybe you can solve it.
Please tell me how to quickly level up deep understandings and analysis and applications of Abstractions.
Once you or I figure it out, I’ll write Part 2.
Author’s note: I especially welcome your comments on this article. I truly desire your help in figuring this out.
Professor Samuel Ichiyé Hayakawa in his seminal work Language in thought and action discusses how abstraction is crucial to our use of language. This built off of earlier work done by Alfred Habdank Skarbek Korzybski who developed the field of general semantics. In his book Science and Sanity: An Introduction to Non-Aristotelian Systems and General Semantics, he introduced the concept of the Structural Differential (an overview and videos of Korzybski introducing the concept is available here.) This concept showed how the brain used refinement via abstraction to make sense of the world on a moment by moment basis.
Abstractions are models of reality not reality itself. As you climb the levels of abstraction be sure to periodically slide down the layers of abstraction. Navigating the lattice of abstraction is a skill and being specific is a practice that help to keep your abstractions grounded in reality. All abstractions will lose detail, that is their whole point, so it is important to test your abstractions with concrete examples to make so that the abstractions still encompasses all relevant details.